Skip to main content
VFACE

Attendance Intelligence

Attendance without the guesswork.

SPACES fuses HD video, thermal imaging, and edge AI to deliver 99.2% first-pass accuracy — eliminating proxy attendance across every campus and shift.

Request a DemoSee plans
99.2%
first-pass attendance accuracy
<500ms
edge decision latency
0
false negatives in a 6-month trial
<30s
first-period start delay, from 4 minutes

What Spaces delivers

99.2% first-pass recognition accuracy in live deployments
Multi-modal fusion: HD video + thermal + edge AI
Zero proxy attendance — photographic audit trail per entry
Real-time administrator dashboard with live headcount
Scales to 10,000+ users per campus with no added staff overhead

In production

Proven at VFACE Matric School, Tiruchirappalli

Running across 500+ students as the school’s system of record for attendance.

  • 99.2% first-pass accuracy
  • 0 false negatives over a 6-month operational trial
  • First-period start delay cut from 4 minutes to under 30 seconds
  • Manual roll-call retired

Features

Everything the platform does, by area

Capture

Multiple sensors, one identity decision.

  • HD video face recognition
  • IR thermal for low light and fever flags
  • Audio cues for classroom presence

Reconcile

Conflicts are resolved by the system, not by teachers.

  • Bayesian multi-modal fusion
  • Identity confidence scoring
  • Human review queue for ambiguous cases

Act

Presence turns into decisions.

  • Live teacher and operations dashboards
  • Parent attendance confirmations on mobile
  • Truancy and pattern alerts

Prove

Ready for regulators and disputes.

  • Sensor-level audit trail
  • DPDPA-aligned retention and purging
  • On-premises option for compliance-critical sites
  • Multiple sensors, one identity decision.

  • HD video face recognition
  • IR thermal for low light and fever flags
  • Audio cues for classroom presence
  • Edge AI at the door, no cloud dependency

How it works

From first call to live in four steps

01

Survey and place sensors

We map entry points and classrooms, then size the video, thermal and audio array for each.

02

Enrol with consent

Students and staff are enrolled with verifiable consent. Templates stay on your infrastructure.

03

Recognise at the edge

On-device models decide in under half a second, so doors never wait on the internet.

04

Reconcile and report

The fusion engine writes one record per person, flags doubt, and updates dashboards and parents.

Connects with your stack

  • SchoolOne
  • HRMS
  • Sage
  • Tally
  • Razorpay
  • Brevo SMS
  • Flutter mobile
  • PostgreSQL
  • TimescaleDB
  • YOLOv8
  • NVIDIA Jetson
  • REST and webhooks
  • SchoolOne
  • HRMS
  • Sage
  • Tally
  • Razorpay
  • Brevo SMS
  • Flutter mobile
  • PostgreSQL
  • TimescaleDB
  • YOLOv8
  • NVIDIA Jetson
  • REST and webhooks

The challenge

Roll-call, swipe cards and fingerprints no longer give you a trustworthy record

Manual attendance does not scale

At 500+ students, roll-call eats the first minutes of every period and still leaves gaps and proxy marks.

Single sensors fail quietly

Fingerprints fail on damp hands and cards get lent. One failed modality should not become a wrong record.

Multi-site blind spots

Leaders of several campuses see yesterday’s numbers, if any. Presence needs to be one live source of truth.

How we build

Engineering principles behind every engagement

Edge first, cloud second

Doors must work when the network does not. Core recognition runs on the device.

  • Sub-500 ms on-device inference
  • Cloud re-identification only for low confidence
  • Offline buffering and resync

Technology we work with

  • YOLOv8
  • NVIDIA Jetson
  • PostgreSQL
  • TimescaleDB
  • Next.js
  • Flutter
  • Bayesian fusion
  • ONNX Runtime
  • Docker
  • Redis
  • YOLOv8
  • NVIDIA Jetson
  • PostgreSQL
  • TimescaleDB
  • Next.js
  • Flutter
  • Bayesian fusion
  • ONNX Runtime
  • Docker
  • Redis

Your first 90 days

What you can expect, phase by phase

Days 1–30

Survey and pilot

  • Site survey and sensor plan
  • Consent and DPDPA workflow
  • Enrolment of pilot classes
  • Baseline of manual accuracy
Days 31–60

Deploy and tune

  • Sensors installed at entries
  • Shadow run beside roll-call
  • Threshold tuning on live data
  • Teacher and parent onboarding
Days 61–90

Cut over and scale

  • Roll-call retired
  • Audit and review procedures
  • Reports for leadership
  • Plan for remaining campuses

Market comparison

How it compares with hardware-first biometric and face-attendance vendors

Comparison with Hardware-first biometric and face-attendance vendors
CriterionVFACE This productHardware-first biometric and face-attendance vendors
Identity methodVideo + thermal + audio fusedNot publicly stated (Truein markets face recognition)
Decision locationEdge first, cloud fallbackNot publicly stated
Ambiguity handlingConfidence score and review queueNot publicly stated
School contextParents, truancy, multi-campus built inBuilt for workforce and multi-site staff attendance
Audit trailSensor-level, DPDPA-alignedNot publicly stated
Book a walkthrough

Example: Truein (AI face-recognition attendance for multi-site workforces; truein.com/pricing lists 70+ attendance policies and no public price, checked 2026-10-09). Based on public vendor information at the time of writing; capabilities and pricing change, so verify against current vendor pages.

Industry benchmark

Attendance programmes: SPACES vs typical

First-pass accuracy

VFACE programmes99.2%
Industry average94%

Time to mark a class

VFACE programmes30 s
Industry average240 s

Lower is better

Proxy marks caught

VFACE programmes98%
Industry average60%

Figures are indicative ranges from typical engagements compared with published industry surveys. Results vary by scope and starting point.

Plans

Per student, per month, plus a one-time sensor bill

Campus

Quoted

per student / month

  • Single-site deployment
  • Dashboards and parent alerts
  • Standard support
Talk to us

Multi-campusRecommended

Quoted

volume pricing

  • Cross-site presence view
  • HRMS and SchoolOne integration
  • Priority support
Talk to us

On-premises

Quoted

annual licence

  • Fully on your servers
  • Compliance-critical sites
  • Dedicated engineer
Talk to us

Typical payback is 8 to 12 months through staff time saved alone. Final pricing depends on site count and sensor design.

Client voices

What our clients say

4 min to 30 s

first-period start delay

“The first period now starts when the bell rings. We stopped arguing about who was present and started acting on who was not.”
R. MeenakshiVice Principal, VFACE Matric School

Security and governance

Built for children’s data

DPDPA aligned

Verifiable consent, purpose limitation and automatic purging.

Audit trail

Every decision keeps its sensor evidence for investigation.

On-premises option

Templates and footage can stay entirely on your servers.

Access control

Role-based views; footage access is logged.

Frequently Asked Questions

Ready to see Spaces in action?

Our team will walk you through a live demo in your context — no sales pitch, just the product doing its job.